fasm doesn't handle repe cmps instruction?

Index > Main > fasm doesn't handle repe cmps instruction?

Author

Thread

AlexP

Joined: 14 Nov 2007
Posts: 561
Location: Out the window. Yes, that one.

AlexP 24 Nov 2007, 02:37

After trying to get a loop working for validating passwords, I tried using a new instruction. The repe cmps instruction (from intel manual) doesn't seem to work with fasm, it just gives me a "invalid address" error. The variable FileHeaderBuffer points to a 64-byte long buffer for a header from a file passed to the function, but the "invalid address" is telling otherwise. Why is this?

The buffer code in the .data section looks like this:

FileHeaderBuffer RB 64

Code:


                PasswordCheck:

                ;Get offset of hashed password in file                
                ;-- -----------------------------------
                ;mov pointer to password in header into eax
                lea eax,[FileHeaderBuffer+0x8] ;offset
                ;pointer to user-supplied hashed pass already in ebx
                ;-------------------------------------


                ;Save registers for repeat instruction
                ;-----------------------------------
                ;Store the two pointers in temp regs
                push edi
                push esi
                ;Operands need to be in EDI and ESI
                mov edi,eax
                mov esi,ebx
                ;-----------------------------


                ;Repeatedly compare strings
                ;------------------------------
                ;Repeat instruction counter is ecx
                xor ecx,ecx  ;WOW ALMOST BIG BUG!!!!
                mov ecx,4
                ;This instruction compares what's in edi and esi
                ;then repeats for ecx times
                ;In other words, compares two 16-byte values
                ;  which in this case are our encrypted passwords
                repe cmps [edi],[esi]
               ;------------------------------
                je somewhereelse....

The instruction "repeats a string instruction the number of times specified in the count register or until the indicated condition of the ZF flag is no longer met.". I have seen it work in code before, but it is not working for me.

24 Nov 2007, 02:37

LocoDelAssembly
Your code has a bug

Joined: 06 May 2005
Posts: 4624
Location: Argentina

LocoDelAssembly 24 Nov 2007, 03:17

Because you can only use ESI as dest and EDI as source.

TFM wrote:

cmps subtracts the destination string element from the source string element and updates the flags AF, SF, PF, CF and OF, but it does not change any of the compared elements. If the string elements are equal, ZF is set, otherwise it is cleared. The first operand for this instruction should be the source string element addressed by SI or ESI with any segment prefix, the second operand should be the destination string element addressed by DI or EDI.

cmpsb ; compare bytes
cmps word [ds:si],[es:di] ; compare words
cmps dword [fs:esi],[edi] ; compare double words

In your case rep cmps word [esi], [edi] or just rep cmpsw will suffice

24 Nov 2007, 03:17

AlexP

Joined: 14 Nov 2007
Posts: 561
Location: Out the window. Yes, that one.

AlexP 24 Nov 2007, 03:19

Yes, but I wanted to use repe (repeat if equal) so that if it breaks and the zero flag is not set, then I know the four dwords are not equal. So should I just use rep cmps dword [esi],[edi]? By the way, it is still giving me an invalid address error.. I'll check my code over, but it shouldn't be saying that..

24 Nov 2007, 03:19

LocoDelAssembly
Your code has a bug

Joined: 06 May 2005
Posts: 4624
Location: Argentina

LocoDelAssembly 24 Nov 2007, 03:47

I use "repe cmps dword [esi],[edi]" here and works OK, perhaps the error is somewhere else (also)?

Code:

virtual at 0
  rep
  repe
  load a byte from $-2
  load b byte from $-1
  if a = b
    display "same opcode!!", 13, 10
  end if
end virtual

; Guess what it prints

Sorry though, it was a misspelling using REP instead of REPE and after reading I couldn't see it because I saw that it was correctly written for the purpose you wanted even though it was a little "obfuscated".

24 Nov 2007, 03:47

AlexP

Joined: 14 Nov 2007
Posts: 561
Location: Out the window. Yes, that one.

AlexP 24 Nov 2007, 04:01

...Okay.. thanks for the little opcode demonstration, I think it's in my memory handling. I'll try to find out what is causing the "invalid address" message. Fasm error handling should be able to point out which address or specific opcode has the issue in the future. THanks!

24 Nov 2007, 04:01

LocoDelAssembly
Your code has a bug

Joined: 06 May 2005
Posts: 4624
Location: Argentina

LocoDelAssembly 24 Nov 2007, 04:31

But it does! It tell you which line of code is problematic, at least with FASMW.

24 Nov 2007, 04:31

AlexP

Joined: 14 Nov 2007
Posts: 561
Location: Out the window. Yes, that one.

AlexP 24 Nov 2007, 17:11

Yeah, which line of code, but I'm saying which operand has the problem.. Or better yet, fasm should print out a paragraph saying what the problem was, where it happened, what I could've done to fix it, and whether to have one or two sugars in my morning coffee lol... I'm just gonna have to set the file pointer and read in those bytes manually instead of from the GIANT data structure I had. Great....

24 Nov 2007, 17:11

< Last Thread | Next Thread >

Forum Rules:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum